P-262 Comparative ability of iDAScore and embryo morphology grade to predict clinical pregnancy: Retrospective cohort study of 1510 fresh and frozen elective single embryo transfers

Abstract Study question Can iDAScore predict ongoing clinical pregnancy (CP) with a sensitivity and specificity equivalent to that associated manual morphology assessment grading? Summary answer The fully automated was able CP performance grading in this retrospective cohort study. What is known already an artificial intelligence (AI) based algorithm developed by applying machine learning morphokinetic time laps (TL) image data of embryos treatment outcome. Embryos are automatically assessed on day 5 culture ranked according iDAScore, ranging from 1 9.9. may be prioritised for transfer highest sore. Along other published AI algorithms, has been proposed optimise the chance following ET improving objectivity embryo compared scoring systems. autonomous algorithms beneficial implications laboratory workload. design, size, duration Retrospective audit 787 fresh 723 frozen single cycles which took place April 2019 September 2022. All recipient, surrogacy, warmed oocyte, biopsy, cleavage stage slow thaw were excluded. Participants/materials, setting, methods Selection blastocyst grade. iDAScores obtained retrospectively. area under receiver operator characteristic (AUROC) curve prediction both grade overall two stratified analyses. first separately, second their female age groups ≤ 35 >35 years. defined ultrasound detection foetal heartbeat. Main results role mean blastocysts iDA 8.34 ± 1.4, strong correlation classic morphological (r = 0.69, P < 0.001). rate 31.0% (95%CI 27.9-34.3) 44.0% 40.8-48.0) respectively. score positively rates (adjOR 1.69, 95%CI 1.44-1.99) transfers 1.45, 1.26-1.67), independent maternal age. There no difference AUROC (AUC 0.64, 0.62-0.67) versus conventional (AUROC 0.63, 95CI 0.61-0.66) when all eSETs considered, or eSETS 0.66 0.62-0.70 vs 0.65 0.62-0.69) 0.63 0.59-0.67 0.61 0.57-0.64) considered separately. exhibited slightly better analyses, higher women years; 0.68 0.64-0.71 0.64 0.60-0.67, p 0.021) but observed younger women. This primarily driven (p 0.002). For >35years 8.75 67% 60% pregnancy. Limitations, reasons caution centre study needs confirmed prospectively multi-centre trial. That selection morphology, have contributed overestimation model performance. Wider findings application ranking transfer, potential maintain while reducing burden / assessments. Trial registration number Nil